Effects of child characteristics on ADI-R: Electronic Supplement1

Collaborative Programs of Excellence in Autism (CPEA) Diagnosis

CPEA Diagnoses of Autism, Asperger Disorder or ASD are determined hierarchically and based upon scores from the ADI-R, ADOS, and the Best Estimate Clinical (BEC) diagnosis made by the supervising clinicians; criteria for Aspergers also include consideration of child age, IQ and language milestones. These criteria were established by the CPEA to ensure standardized diagnostic classification across sites and adopted by the SSC under the same rationale. The criteria used in the SSC were slightly modified from those described by Lainhart et al. (2006):

  1. For a CPEA diagnosis of “Autism” an individual must meet the following:
  2. ADI-R classification of “Autism.” This is based on meeting published cut-offs on the ADI-R diagnostic algorithm (LeCouteur et al., 2003) in the domains of Reciprocal Social Interaction (SOC), Communication, Restricted, Repetitive & Stereotyped Patterns (RRB) of Behaviors, and Age of Onset.
  3. ADOS classification of “Autism” or “Autism Spectrum.” This is based on meeting published cut-offs on the revised diagnostic algorithms for Modules 1-3 (Gotham et al., 2009) and cut-offs on the originally published diagnostic algorithms for Module 4 (Lord et al., 2000).
  4. BEC Diagnosis of “Autism,” “Autism Spectrum” or “Aspergers.”
  1. For a CPEA diagnosis of “Aspergers” an individual must not meet criteria for “Autism” and must also meet the following:
  2. Chronological age of 5 years or older
  3. Verbal IQ of 80 or above
  4. Age of First Words (from the ADI-R) is 24 months or younger
  5. Age of First Phrases (from the ADI-R) is 33 months or younger
  6. ADI-R classification is not “Autism”
  7. ADI-R SOC domain score is 10 or higher
  8. ADI-R RRB domain score is 2 or higher
  9. ADOS classification of “Autism” or “Autism Spectrum” or ADOS Social+Communication Total (based on originally published algorithms) is 4 or higher
  10. BEC Diagnosis of “Autism,” “Autism Spectrum” or “Aspergers.”
  1. For a CPEA diagnosis of “Autism Spectrum Disorder” an individual must not previous criteria for “Autism” or “Aspergers” and must also meet the following:
  2. ADI-R classification of “Autism Spectrum.” This is based on CPEA criteria (Lainhart et al., 2006; Risi et al., 2006), which requires one of the following:
  3. Meeting cut-offs on the SOCand Communication domains
  4. Meeting cut-offs on either the SOCor Communication domain and score within 2 points of the cut-off on the other
  5. Score within 1 point on both the SOCand Communication domains.
  6. ADOS classification of “Autism Spectrum.” This is based on revised diagnostic algorithms for Modules 1-3 and originally published diagnostic algorithms for Module 4.
  7. BEC Diagnosis of “Autism,” “Autism Spectrum” or “Aspergers.”
  1. If none of the above criteria are met, the participant receives a CPEA diagnosis of “Nonspectrum”

Effects of child characteristics on ADI-R: Electronic Supplement1

Table S1
Predictors of ADI-Diagnostic domain totals
SOC+NVC / RRB
B / SE B / 95% CI / rpart / R2 / ΔR2 / f2 / B / SE B / 95% CI / rpart / R2 / ΔR2 / f2
Lower / Upper / Lower / Upper
Constant / 29.72 / .16 / 29.41 / 30.03 / 5.90 / .05 / 5.81 / 5.99
ASD-Symptoms
ADOS-CSS / .82 / .09 / .64 / 1.00 / .16 / .039 / .039 / .041 / .12 / .03 / .06 / .17 / .09 / .008 / .008 / .008
Demographics
Gender / -.81 / .45 / -1.69 / .06 / -.03 / .053 / .013 / .014 / -.45 / .13 / -.70 / -.20 / -.07 / .014 / .006 / .006
Race / .53 / .37 / -.20 / 1.27 / .03 / -.24 / .11 / -.45 / -.03 / -.05
Mat Educ / .60 / .31 / -.01 / 1.21 / .03 / .01 / .09 / -.17 / .18 / .00
Behavior Problems / .053 / .000 / .000 / .032 / .018 / .019
CBCL-E / .01 / .02 / -.02 / .04 / .01 / .01 / .00 / .00 / .02 / .04
CBCL-I / .05 / .02 / .01 / .09 / .04 / .03 / .01 / .01 / .04 / .09
Developmental Level / .292 / .239 / .338 / .046 / .014 / .015
Age / .74 / .05 / .64 / .83 / .25 / .04 / .01 / .01 / .07 / .06
M1vsM3 / 7.87 / .58 / 6.74 / 9.01 / .24 / .35 / .17 / .02 / .67 / .04
M2vsM3 / 3.61 / .45 / 2.72 / 4.49 / .14 / .26 / .13 / .01 / .52 / .04
Nonverbal IQ / -.03 / .01 / -.05 / -.02 / -.07 / .00 / .00 / -.01 / .00 / .00
M1vsM3*age / .19 / .12 / -.05 / .43 / .03 / .14 / .03 / .07 / .21 / .08
M2vsM3*age / .91 / .12 / .67 / 1.15 / .13 / .06 / .03 / -.01 / .13 / .04
Note: Soc+NVC= Reciprocal Social + Nonverbal Communication domain totals; RRB=Restricted, Repetitive and Stereotyped Behavior domain total; Mat Educ=Maternal Education; ADOS CSS=Autism Diagnostic Observation Schedule Calibrated Severity Score CBCL=Child Behavior Checklist; I=Internalizing; E=Externalizing; M1=ADOS Module 1; M2=ADOS Module 2; M3=ADOS Module 3

When ADOS-CSS was entered last into the models predicting diagnostic SOC+NVC and RRB totals, effect sizes for each set of predictors remained highly similar to the model shown in Table S1 (SOC+NVC: demographics f2=.013, behavior problems f2=.000, developmental level f2=.347 and ADOS-CSS f2=.035; RRB: demographics f2=.006, behavior problems f2=.017, developmental level f2=.017 and ADOS-CSS f2=.008).

Table S2
Predictors of ADI-Current domain totals
SOC+NVC / RRB
B / SE B / 95% CI / rpart / R2 / ΔR2 / f2 / B / SE B / 95% CI / rpart / R2 / ΔR2 / f2
Lower / Upper / Lower / Upper
Constant / 11.47 / .10 / 11.27 / 11.66 / 4.66 / .04 / 4.58 / 4.74
ASD-Symptoms / .022 / .022 / .022 / .013 / .013 / .013
ADOS-CSS / .50 / .06 / .38 / .62 / .14 / .14 / .03 / .09 / .19 / .11
Demographics
Gender / .37 / .30 / -.22 / .96 / .02 / .032 / .010 / .010 / -.33 / .13 / -.59 / -.08 / -.05 / .017 / .004 / .004
Race / -.40 / .25 / -.89 / .09 / -.03 / -.29 / .11 / -.50 / -.08 / -.05
Mat Educ / .08 / .21 / -.33 / .49 / .01 / .05 / .09 / -.12 / .23 / .01
Behavior Problems / .055 / .022 / .023 / .057 / .039 / .041
CBCL-E / .04 / .01 / .02 / .06 / .06 / .02 / .00 / .01 / .03 / .09
CBCL-I / .09 / .01 / .07 / .12 / .12 / .03 / .01 / .02 / .04 / .11
Developmental Level / .335 / .280 / .421 / .084 / .027 / .029
Age / .10 / .03 / .03 / .17 / .05 / .01 / .01 / -.02 / .03 / .01
M1vsM3 / 8.42 / .39 / 7.66 / 9.17 / .37 / .60 / .16 / .28 / .93 / .07
M2vsM3 / 2.58 / .29 / 2.02 / 3.15 / .15 / .23 / .12 / -.01 / .47 / .04
Nonverbal IQ / -.01 / .01 / -.02 / .01 / -.02 / -.01 / .00 / -.01 / .00 / -.06
Note:bold = p<.001; italics = p<.05; Mat Educ=Maternal Education; ADOS CSS=Autism Diagnostic Observation Schedule Calibrated Severity Score CBCL=Child Behavior Checklist; I=Internalizing; E=Externalizing; M1=ADOS Module 1; M2=ADOS Module 2; M3=ADOS Module 3

When ADOS-CSS was entered last into the modelscurrentdiagnostic SOC+NVC and RRB totals, effect sizes for each set of predictors remained highly similar to the model shown in Table S2(SOC+NVC: demographics f2=.010, behavior problems f2=.022, developmental level f2=.416 and ADOS-CSS f2=.029; RRB: demographics f2=.004, behavior problems f2=.040, developmental level f2=.033 and ADOS-CSS f2=.012).